David Hume, one of the most prominent members of the Scottish Enlightenment, was a multifaceted Enlightenment thinker. Based in Edinburgh, Hume was a cultivated, cosmopolitan intellectual, a frequenter of Paris salons. Once a diplomat, he wrote a very good history of England. He composed a moral philosophy based in sentiment or feeling, each exhibiting a fine sense of the practical life.

As a nonbeliever, he wrote without any use of God or the transcendent in his philosophical work. The human mind is constituted, for Hume, by nature, and human history is contingent, replete with accidents, not the progressive revelation of a divine theme or developing toward an ideal conclusion. His work regularly ran afoul of religious convention; as a librarian, he once got into trouble for ordering books regarded as indecent. Various writings were temporarily suppressed, and his excommunication was called for. It was a matter of widespread public interest that on his deathbed, Hume did not repent or pray to God. But it’s his epistemology alone that led to truly radical conclusions. The way he did this was to push empiricism far beyond Locke and Berkeley. Hume’s question was, just what does a consistent empiricism commit us to? His answer was very nearly, to no knowledge at all. We’ll examine the key points of his famous Enquiry Concerning Human Understanding, most famously his argument against causality and his legacy.

Hume began his theory of knowledge simply: All knowledge and all ideas derive from experience; this makes him an empiricist. He distinguished impressions and ideas; that is, all the contents of mind for Hume are either impressions or ideas. Impressions are the lively, forceful perceptions that force themselves on us at each moment of experience—in effect, impressions are experience, are all sensations—whereas once those impressions have entered the mind, they begin to decay in forcefulness and become ideas. Ideas are decayed impressions, less forceful impressions. Thus, for example, the impression of my clapping my hands: When I clap my hands, you hear a sound; but now you can remember the sound and remember the clap. The original experience is an impression; the memory is an idea. The human mind can then remember, combine, split apart, change, and analyze these ideas. This is already a clue as to how radical Hume is going to get, by the way, for notice the only distinction between a perception and a memory or imagination is vivacity or force—a quantitative distinction. The implication is that if I took one of my memories, made it more and more vivacious or more forceful, I wouldn’t be able to distinguish it from a perception, from an impression. Gone from this are all Lockean notions that some ideas are felt entering the mind—in other words, that you can distinguish an idea that’s caused by a physical object in the world from one that isn’t.

Now comes Hume’s most basic critical tool, which he refined from a distinction made by Leibniz. Two, and only two, categories exhaust all knowledge for Hume: knowledge of relations of ideas on the one hand and knowledge of matters of fact on the other. Relations of ideas are just what they sound like: They are the relations one idea bears to another in my mind. Relations of ideas are made true logically by those relations, for example, among the subject and predicate of a sentence: All bachelors are unmarried, or two plus two equals four. Let’s take “all bachelors are unmarried.” The statement is true; in fact, it has to be true. Why? Because since you know the meaning of the word “bachelor”—and once you know the meaning of the word “unmarried,” you know that bachelors have to be unmarried— the statement is made true by the relations of the ideas in the statement. Matters of fact are different: They are made true by a statement’s relation to impressions that confirm it. For example, if I say not, “All bachelors are married,” but if I say, “There are bachelors in this room,” or if I say, “The world contains bachelors”; the only way to know that’s true is to go and look, or ask someone to go and look. In other words, someone has to have a series of impressions that confirm that it’s a matter of fact that the ideas in the sentence conform to the world.

Note that both of these are based in experience; don’t get confused and think that relations of ideas somehow don’t come from experience. In matters of fact, a concept previously gotten from experience—I have an impression; it decays into an idea—I now compare that idea to some new impression. For example, at one point I see a swan. I have the idea of “swan” in my head; at some future date I see a swan and say, oh, there is a swan in that lake. It’s true if there is indeed a swan in that lake. On the other hand, when I relate two ideas in my mind, each of those ideas have been gotten from impressions, but now I’m comparing them together. For example, when I say, “Venus is the morning star,” I’m realizing that the word “Venus” and the words “morning star” refer to the same thing.

OK, so what’s the problem? The problem is that each type of knowledge—and we only have two types now, relations of ideas and matters of fact—brings with it good news and bad news. The good news about relations of ideas is that they are absolutely certain; they are universally and necessarily true. All bachelors are unmarried. That’s true in China, the United States, anywhere. It also must be true; it’s necessarily true; it cannot be false. As Hume himself says, its contrary would be a contradiction. If I were to say, for example, “Joe is a bachelor, and yet Joe is married,” I’ve made a contradiction; so that can’t be true. But there is a problem with relations of ideas: They are true only by definition; in effect, they are trivial. They merely tell us that one idea is related to another. If you know what a bachelor is, you already know that bachelors are unmarried. So if someone says to you, “All bachelors are unmarried,” all they’re doing is telling you what you already know. On the other hand, there are matters of fact: Matters of fact give us real knowledge about the real world, that is, about impressions, about experience. But matters of fact are never certain; they are never true by necessity. They tell us what is contingently true, not what must be true.

Hume applied this epistemic dualism to everything, but most famously to causality. Here is how he did it: First, Hume asserts that we have two notions of causality: constant conjunction and necessary connection. Constant conjunction simply means that whenever A has occurred, B has occurred; whenever I’ve seen smoke, I’ve seen fire. To state that is to state a case of constant conjunction. Necessary connection, on the other hand, says more: If I say A and B are necessarily connected, then I say not merely whenever A has occurred, B has occurred, but I say something more. Whenever A occurs, B must occur; or, what is the same thing for Hume, whenever A occurs, A has the power to make B occur; or, what is also the same thing for Hume, whenever A occurs in the future, B will occur. All these three ideas—the will, the must, and the notion of power, that one event can make another event happen, ascribe necessity to the connections between events.

Hume applies his epistemic dualism to necessary connection. Remember, all knowledge is either matters of fact or relations of ideas. Here’s the question: Is there necessary connection in the real world? There’s certainly constant conjunction; there are lots of events where whenever I see A, I see it followed by B. We see the cue ball strike the eight ball; the eight ball moves. Perhaps I’ve seen that every day of my life playing pool. But I do not see the necessity of its moving; in other words, we have no experience of the “have to.” If one says, when the cue ball hits it, the eight ball has to move, you see it move. You don’t see the “has to,” the necessary; you don’t see the power or force by which the first thing makes the other happen. All we have is one experiential snapshot following another: the snapshot of cue ball moving and striking the eight ball, and then the snapshot of the eight ball moving. That’s constant conjunction; to say any more than that, Hume says, is to make a claim not based on experience.

More completely, Hume argued that our general knowledge of matters of fact— and by “general knowledge,” he means not just reports of contemporaneous experience, like “I am right at this moment experiencing the lectern,” but my more general kind of knowledge, the kind of knowledge that science has, which is to say, “Living organisms contain carbon” or something like that—so our general claims about the world make a presupposition; and that presupposition is “The future will resemble the past.” Indeed, all of science makes that presupposition. If it didn’t, then science couldn’t say, “The Sun will rise tomorrow”; it could say, “Force equals mass times acceleration in the past,” but we don’t know if force will equal mass times acceleration in the future. This means it couldn’t utter any general laws regarding matters of fact; so science needs to be able to say that the future in some respects will resemble the past.

Hume’s question is, is it true? Can we know that the future will resemble the past? For it to be true, it must be true either by relations of ideas or matters of fact experience. It’s not true by relations of ideas; the definition of the word “future,” the definition of the word “past,” do not imply any connection or resemblance. There’s nothing about the statement “Whatever happens in the future will resemble what’s happened in the past” that is like “All bachelors are married” or “Two plus two equals four.” If the future does resemble the past, if that’s true, it’s not going to be true by definition; so it’s not relations of ideas. What about the other option; is it true by matters of fact? Here, Hume points out nor can it be true by general facts of experience, by matters of fact, for we have already said that such general facts presuppose that the future will resemble the past. We cannot now use general matters of fact to prove that the future will resemble the past; that would be a logical circle. In effect, what Hume is saying is our claim that the future will resemble the past is, whether true or false, unknowable; there is no way we can know that is true. All our claims of science, that certain things must necessarily happen, are based on that presupposition that we can’t possibly know to be true because it doesn’t fit into either knowledge of the relations of ideas or knowledge of matters of fact.

The general knowledge we gain from experience—for example, the idea that the Sun will rise tomorrow, which all of us believe—presupposes something, namely, that the future will resemble the past. Or what is the same thing, it presupposes necessary connection; it presupposes that when some events happen, other events must happen. All that is a presupposition, but experience can’t validate it. We may put this simply: Neither the future, nor necessity, can be experienced; we have no knowledge of either. We cannot experience the future, and we cannot experience necessity; therefore empiricism, according to Hume, must say that we have to cease to use such concepts. We are left with constant conjunction; necessary connection is out.

It’s crucial to recognize the radical nature of this conclusion, and even some students of philosophy sometimes miss this point. Hume is not merely saying that our knowledge of the future is uncertain; he’s not merely saying it’s merely probably; he’s saying something much worse: He’s saying we have no probable knowledge of future either. We have no reason to believe that the coin will drop; no reason to believe that the Sun will rise tomorrow; no reason to believe that if I jump off the building tomorrow, I will fall. None; no reason whatsoever; no rational justification for such beliefs. This means, by the way, that the very method of induction is in peril. Induction—we haven’t spoken of this before, but—one of the methods by which science works and by which we make inferences and try to gain knowledge is induction. Induction is the method by which we infer a probable truth about a population of events from observing a selected sample of events. What Hume is claiming is the method induction has no rational justification.

Let’s take an example: When we judge from selecting 10 brown beans from a kettle—we have a kettle; we have a lot of beans in there—we select 10 beans, look at their color, and they happen to be brown. Now we’re asked the question, make a guess about the color of the other beans in the kettle that you can’t see and haven’t picked yet. We’re going to suggest that the beans in the kettle probably are all brown, because these 10 are brown. This is the same reasoning that we use when we use a key to open a lock, and it’s worked 10 times in a row, and we believe that the key will work the next time. Induction only implies probable truth; nobody says it’s absolutely certain the key will work the next time, because we’ve all had experience with keys that don’t work, and sometimes the sample you select from a population may be biased and not represent the whole population. Nevertheless, the way our minds work is to make inductive inferences from samples to population. But Hume’s point is that there’s no rational reason to believe that at all; there’s no justification of induction.

This is pretty bad—but Hume isn’t finished. Since we have no knowledge, not even probable knowledge, of necessity in events, the future, we also can’t know the existence of forces that are unexperienced, like gravity. For Hume, there is absolutely no evidence that gravity exists. One immediately wants to respond to this, this is crazy; are you saying if I jump up, I won’t fall down? If I throw a ball, it won’t fall down? Hume says, experience gives us evidence that things fall, but it doesn’t give us evidence that there’s an unseen force making them fall. To believe in gravity is to believe in something, and what kind of thing is it? A force, a power, that is itself unseen and makes other things happen. Hume says, if you’re going to say that’s true, you might as well imagine that there are little gnomes, or little sprites or spirits, in the world making things happen. In other words, from Hume’s point of view, it’s completely unempirical, even unscientific, to go beyond the claim “things fall” to the claim “something unseen makes them fall.”

Remember as well that when we looked at Locke, we saw Locke recognized in each material substance not only its experienced qualities—say that the lectern is brown; it’s this tall; it’s solid; it has a whole bunch of properties that I can experience—but Locke had also said that there must be more to the lectern than merely the properties I experience. There must be an unknown support, in Aristotle’s Greek, a hypokeimenon, an underlying something or other—and that constitutes its nature as a material substance. In other words, Locke held it couldn’t be the case that a material substance is just the qualities I perceive—there is some material there, whether I perceive it or not, that’s holding it all together. What is Hume going to do to this idea? It’s not hard to figure. Locke himself had said the unknown support can’t be experienced, but it must be there; and he himself thought, because he accepted the corpuscular hypothesis that many did in the 17th century, Locke thought that the underlying, unknown support was, in fact, little atoms that are too small for us to perceive. But what Hume has to say now, of course, is there is no underlying support. From an empiricist point of view, an object is its perceptible qualities, nothing more; but what this is tantamount to saying is that there is no material substance. There are experienced qualities; the lectern is a list of properties I can experience, and there is nothing more to it.

This applies even further, because it applies to mental substance as well. Berkeley, as we saw, denied material substance, claiming that all experienced qualities inhere only in our own mental substance. But Hume applies the same argument to our minds and our selves: What am I? What is Cahoone’s self? I am the list of memories, impressions, and ideas of which I am aware. I can scour my memory, try to think of everything I’ve ever thought. I’ll be unsuccessful, but as much as I can think of, that list—or as he literally called it, a bundle, or heap, of ideas—is my self. There is nothing beneath that—no underlying mental support or soul substance that supports and unifies. There is nothing to reality, in effect, but impressions, surfaces, that which we perceive. This literally means, for Hume, that we do not know if an object continues to exist when it’s not being experienced. Hume has been led to pure phenomenalism: All that exists is phenomena. At this moment, I have no reason, no rational reason, to claim that the lectern exists, because I am not experiencing it. Now, I have plenty of reasons; I have impressions—the lectern is a matter of fact. But now, I don’t. The fact that he denies material and mental substance leaves him in the position of saying when there’s no perceptible qualities, there’s nothing there. For Locke and Berkeley, they believed there was something there even when there were no perceptible qualities.

As you might imagine, Hume rejects all arguments for the existence of God. We don’t have the time to go into much detail here, but let’s just say, for example, one of the most famous arguments for God had come from the great medieval Scholasticist Thomas Aquinas. It’s one that’s very familiar to all of us: It’s that God is the necessary first cause of the universe. Everything in the universe has a cause: The cause of me is my parents; the cause of my parents are their parents. You keep going back, and at some point you get to the origin of the universe, maybe the Big Bang, and you say, there must be a cause of that; God is the uncaused first cause. But what Hume says in response is that—he makes an interesting claim—he says causality, which remember for him is just constant conjunction, is a relation between events inside nature. Here’s nature. Inside it, there’s a relation between pairs of events called causality. It’s perfectly acceptable to ask about the cause of events in nature; but it’s a false analogy, an illegitimate extension of the concept, to ask what the cause of all of nature would be. We can’t take the relation of causality and go outside the whole of nature and try to apply it to nature itself.

One other argument of the existence of God was the famous teleological, or design, argument. The most famous version of that was given from William Paley early in the 19th century: “Walking on the heath,” he says, “I stub my toe on a pocket watch.” What’s more rational: to believe that the pocket watch arose in this place by random interactions of matter—in other words, the lightning struck on the sands and the mud, and the different minerals in the soil gradually formed stainless steel and glass and little hands turning around and a pocket watch—or is it more rational to believe that some intelligent creature made this watch and left it here? Paley says obviously it’s more rational to believe in a designer; and if this universe is far more complex and evidently contrived than the watch, even more reason to believe the universe has a designer. Hume’s response to this is that if order in the world justifies the claim of a designer—it’s very interesting, he doesn’t take the argument head on; he says, OK, fine, if you want to say whenever there’s order there must be a designer, fine—however, Hume says, the degree of order in this world is not perfect. The word isn’t completely orderly—there are tsunamis, there is hunger, there is poverty, there is misery, planets collide, comets collide. The world is only partly ordered; and if so, then your argument for a creator and a designer can only say that the designer is limited in intelligence and in power. In other words, a partly orderly world can only prove a partly intelligent and powerful God.

Hume, at the end of his inquiry, makes a kind of shocking statement. He says:

When we run over libraries, persuaded of these principles, what havoc must we make? If we take in our hand any volume—of divinity or school metaphysics [which means Scholasticism, the traditional medieval Scholasticism, which still dominated in the universities of Europe even at this point in the middle of the 18th century], for instance—let us ask, Does it contain any abstract reasoning concerning quantity or number [in other words, is it relations of ideas]? No [suppose the answer is no]. [Then we ask] Does it contain any experimental reasoning concerning matter of fact and existence? [Is it matters of fact?] [Suppose the answer is] No. [Then] Commit it ... to the flames, for it can contain nothing but sophistry and illusion.

How can all this be? We all believe that the Sun will rise tomorrow; we all believe that a dropped coin is going to fall; we all believe that if we stand out in front of a truck rushing down the street, bad things are likely to happen. Does Hume pretend not to believe this? No, Hume is not a fool; he admits it. Of course he believes that the Sun will rise tomorrow. We all believe it, but that doesn’t mean it’s rational or right. Hume says that we human beings are designed by nature. Nature has put us together in such a way that whatever happens repeatedly causes our mind to expect it to happen again; this he calls custom, or habit. Custom, or habit, guides our beliefs. We couldn’t live without it, but there is no reason or rationality behind it all. Hume felt this rather deeply. In fact, if I may quote another brief passage, he famously wrote:

Most fortunately it happens, that since [human] reason is incapable of dispelling these clouds [he means clouds of doubt], nature herself suffices ... I dine, I play a game of back-gammon, I converse, and am merry with my friends: and [when] after three of four hours’ amusement, I wou’d return to the speculations, they appear so cold, and strain’d, and ridiculous, that I cannot find in my heart to enter into them any farther. He concludes from this that there is nothing to fear from his skepticism, since “Nature is always too strong for principle.”

What, then, has he left us with? What Hume leaves us with is that reason and life are tragically at odds; this is almost a bit of existentialism in the 18th century. Hume shows that if we begin with empiricism—which is the most reasonable basis, he thinks, for the practice of science and for the wisdom of common life—and a belief that all knowledge comes from sense experience, we in fact must, in the end, come to the conclusion that all our beliefs about the world, all our claimed knowledge about the world, with very few exceptions, have no rational justification whatsoever. Nevertheless, we are not at liberty to cease to believe in them, because nature is too strong for principle.

At any rate, Hume thus provided the model skeptic that all other philosophers would have to try to defeat. Philosophers to this day argue with Hume. His thought led to the greatest of the 18th-century philosophers, and the next chapter in our tale: the work of Immanuel Kant.